Date: 2026-09-01 20:47 Summary: Removed the ~10-row cap on keyword-daily-pool; batch conclusions may all enter the pool. Visibility: people
[question] 剔除词池约 10 行 的限制
[try to solve]
Context
Playbook and funnel docs capped each batch at ~10 rows in keyword-daily-pool.tsv. The user asked to drop that limit.
Process
Updated active rules so pool append has no row-count ceiling: keyword-planner-analysis-rules.md, Skill keyword-to-tool-funnel, ops/seo/keyword-to-tool-ops.md, keyword-to-tool-funnel.md, long-tail strategy §4.2/§5, tracker checklist, bing_serp Next-steps template, prompt-builder analysis wording, and a related reviews note. Tracker decision log records the change.
Root cause / analysis
The ~10 figure was a discovery cadence hint that agents treated as a hard cap. Cap constrained analysis fidelity; the real safety rails are week build ≤1–2, no doorway, and no long_gap without SERP—not pool length.
Solution
- Pool: all batch conclusions may be appended; identical verdict/tier/placement near-synonyms may merge into one row.
- Unchanged: week build quota, absorb-first, no auto work-tasks.
- Historical SERP markdown Next-steps lines left as-is; new Bing batches get the updated template text.
Notes / boundaries
Do not confuse “unlimited pool rows” with “unlimited new tool URLs.”
[actions]
- Updated playbook, Skill, ops, funnel, strategy, tracker, bing_serp
io_util.py, prompt-builder analysis note